Keira's Bookmark

• Source: Assigning CEFR Ratings to ACTFL Assessments (ACTFL 2016)
• Note that they assign different equivalency levels for receptive/productive skills and I have given an average in the table above.

Official Equivalency Table:

The official claim is that there is a 1:1 match between TOPIK levels 1-6 and CEFR A2-C1 (National Institute for International Education, 2024). However, no equivalency is given on the exam certificate (which would require a detailed research and standard-setting agreement) and there is no source given on the documentation making this claim.

Independent research by Korea-based researchers also claim a 1:1 equivalency between the levels (Kim Chungsook et al. 2025 – the text is in Korean, but there is an English abstract). However, they do also note some key weaknesses of the test:

• There is no compulsory spoken section (a separate spoken exam is now available in some countries, but this is a computer-based test without any in-person interaction).
• The reading/listening sections are entirely multiple-choice and the number of questions and length of said questions is limited, especially as it has to cover 4 levels in 50 questions, in the case of the TOPIK II (Intermediate-Advanced) exam. This limits the rigour with which the test is able to test the upper levels.
• They note that Level 4 is a slightly lower level than B2 (although they later claim equivalency because there is much overlap). I have the power to give a more detailed equivalency, which is why I have split B2 into B2 and B2+ to show this distinction.
• They focus on similarities between level descriptors at the higher levels, but don’t note the differences, of which there are some significant ones. Note in the pictures below that level 6 still uses qualifiers such as “most”, where as C2 does not. This doesn’t mean there is no overlap, but I would argue Level 6 starts from a lower point than C2.

Finally, there is my personal experience. I haven taken the TOPIK exam twice and got a level 6 both times. However, despite achieving an extremely high TOPIK grade initially, when I went to Korea after taking it the first time I was placed in a Level 5 class (so I should a level 4 on the test, by that logic, as I was yet to acquire level 5 skills). This highlights where the test can fail at testing the skills each level requires. That being said, I found the reading and listening sections difficult the first time I took the test and made a lot of guesses.

The second time I took the test, two years later, having graduated level 6 of a Korean language programme in Korea, attended university classes in Korean and worked as a translator for over a year (so my Korean was definitely better), I got a higher score on reading but a lower score on writing. This is partially luck, partially a lack of preparation – I prepared a lot for the first test and not at all for the second.

What I can say, is that I would self assess by Korean at the time of 2023 exam as a C1 (and B2 at the time of the 2021 exam) based on my experience with German, and I found the reading/listening section easy(ish) in 2023 – I didn’t have to guess 99.9% of the time. Therefore, I would be happy saying that level 6 can test C1 if you’re answering because you actually know the answer, rather than just having good exam technique (which will always be the shortcoming of multiple choice tests like this).

Also, not to brag, but my exam technique is quite good so these results are probably skewed slightly and not representative 100% of the average test taker. And as you can see I could still get more points, especially in writing. That being said, my general feeling about Level 6 = C1 seems to be echoed by other learners as well, who are even harsher and say Level 6 = B2 (which I would agree with at the lower end of level 6) (see this reddit thread).

Therefore, I think that Level 6 can test and encompass C2, but I think it encompasses C1 as well: it is the upper band, it is everything from C1 and up. Level 5 is probably closer to that upper B2, maybe the very bottom of C1 – the bit that of B2 that is missing from Level 4, as per the researcher’s conclusions. I’m obviously not a professional and this is largely anecdotal, which is why I have also included the research above, but as this is about the equivalencies I use personally, I will make Level 5 = B2+ and Level 6 = C1 and up.

Hanban, a Chinese government organisation who run the HSK, and the Confucius Institute (who used to be under their leadership but is now technically separate) claim that the HSK 1-6 is equivalent to CEFR A1-C2. This has been thoroughly debunked by every other standards assessing organisation and individual researcher who has cared to research it. For instance, the German Association for Chinese Language Teachers suggests the following equivalencies instead:

Although Hanban haven’t announced that this is wrong, it seems they acknowledge it to as they are now preparing to move from HSK 2.0 to HSK 3.0. Looking at the differences between the two will a) help future proof this page a bit and b) show where the current HSK falls short.

Source: DigMandarin.com

It looks like levels 7-9 will correspond to C1 and C2 (and this is was likely possible under HSK 1.0 which was also harder). In terms of the others – HSK6 will be almost the same level in HSK 3.0 as it is in 2.0, so I would suggest that the highest level won’t change. However, it is clear that the requirements at other levels have become much stricter – this will mean a more gradual progression than the current HSK, but also a bigger base of knowledge needed to pass the lower level exams. If we look at the current HSK – CEFR equivalencies offered by FaCh, then the new HSK 3.0 equivalency could look something like this (of course the exact equivalencies will have to wait until it’s officially released and professionals look at it).

I will say, I took HSK4 and then tested into a B1 class, so I think the German Association might have been a tiny bit harsh, but certainly not by much. I would also say that the HSK does have a compulsory writing section that requires summarisation, which is quite tricky and therefore can probably test a very high B2. Some suggest it touches on the bottom end of C1. For now, until HSK 3.0 is out I will display it as such.

Please do note that HSK 3.0 7-9 exams have already started, but it hasn’t been rolled out to lower levels yet, which is why the table I included above says HSK 2.0 but includes HSK 3.0’s 7-9 levels.

TOCFL offers CEFR equivalencies on their official website which are also printed on the official certificates, and they claim to be based on a series of research carried out in the 2010s. Separate equivalencies are published for each of the four skills (reading, listening, writing, speaking) and each test type.

The most commonly taken TOCFL exam at the moment is the Computerised Adaptive Testing (CAT) test, which tests only listening and reading. These find an equivalency between TOCFL Levels 1-6 and CEFR A1-C2. A similar equivalency is also found in other skills and test types, but it is important to note that it is completely possible to pass the CAT at a C1 level but have lower levels in other skills.

Additionally, the CAT exam is also multiple choice. Although it doesn’t have the issue with limited question number and length like TOPIK as questions are adjusted for the individual learner’s level, it is still possible to “guess” the correct answer. Nevertheless, allowing for this, the actual skill level of the content provided to test each level of the TOCFL seems to be equivalent to CEFR.

Official equivalencies for CEFR are given on JLPT exams, however it should be noted that these are valid for receptive skills only as the JLPT does not test productive skills. It is useful to note that the JLPT only claim equivalency up to level C1, not C2. More details (in English) can be found on the JLPT website, where the table below also comes from. The PDF detailing the research is in Japanese only and my Japanese isn’t good enough to read it. I’ve also never taken the JLPT so I have nothing else to add here!